NHANES III (1988-1994), a nationally representative survey coordinated by the Centers for Disease Control and Prevention’s (CDC’s) National Center for Health Statistics (NCHS), provided the data for this study. De-identified data were retrieved from the CDC’s public repository (https://wwwn.cdc.gov/nchs/nhanes/) following the protocols for the secondary analysis of population-based health surveys. We included adult participants (age ≥ 20 years) from the NHANES III (1988-1994) with complete data on H. pylori IgG serology and hepatic ultrasonography. Participants with missing baseline anthropometric data or incomplete follow-up information in the NHANES III Linked Mortality File were excluded. Mortality outcomes through December 2019 were acquired from the National Death Index (NDI) linkage files (updated May 2022), using probabilistic matching of participant identifiers (https://ftp.cdc.gov/pub/Health_Statistics/NCHS/datalinkage/linked_mortality/). The original NHANES III protocol was approved by the NCHS Research Ethics Review Board, and written informed consent was obtained from all participants. Given that only anonymized, publicly available data were used, an individual ethics review was not required.

Type 2 diabetes mellitus status was ascertained through three complementary approaches: self-reported physician diagnosis, current utilization of insulin or oral hypoglycemic agents, or fulfillment of biochemical thresholds [fasting plasma glucose (FPG) ≥ 7.0 mmol/L, glycosylated hemoglobin (HbA1c) ≥ 6.5%, or 2-h postprandial glucose ≥ 11.0 mmol/L]. Hypertension was characterized by meeting either of two conditions: (1) documented systolic blood pressure ≥ 140 mmHg or diastolic pressure ≥ 90 mmHg; OR (2) a prior clinical diagnosis of hypertension or active use of antihypertensive medications.

Standardized clinical records provided an extensive array of biochemical indices, including complete blood count parameters, C-reactive protein (CRP) levels, glucose metabolism markers (FPG and HbA1c), hepatic transaminases [alanine aminotransferase (ALT) and aspartate aminotransferase (AST)], serum lipid profiles (total cholesterol and triglycerides), and serum creatinine concentrations. All laboratory assays were performed in accordance with rigorously controlled measurement protocols.

Categorical variables were presented as frequencies (percentages), whereas continuous variables were reported as mean values with standard deviations. Group comparisons for normally distributed variables were conducted using independent-samples t-tests, non-normally distributed parameters were analyzed using Mann-Whitney U-tests, and categorical data were assessed using Pearson’s chi-squared tests. The association between H. pylori seropositivity and MAFLD risk was investigated using logistic regression analysis. Kaplan-Meier curves were used to evaluate survival probability differences between H. pylori antibody-positive and antibody-negative groups. The relationship between H. pylori seropositivity and all-cause mortality was examined using the Cox proportional hazards regression model.

To address potential multicollinearity among the variables, we first assessed pairwise collinearity using the variance inflation factor and correlation coefficients. Strong multicollinearity was detected among the following variable pairs: BMI and waist circumference; FPG and HbA1c; ALT and AST; and serum cholesterol and triglycerides. Based on this criterion, waist, HbA1c, AST, and triglycerides were selected over their collinear counterparts (BMI, FPG, ALT, serum cholesterol, respectively) for inclusion in the multivariate Cox proportional hazards regression model. A two-tailed significance threshold of P < 0.05 was applied for all statistical tests. Data analyses were conducted using the R software version 4.3.3 (www.r-project.org). The survival analysis was performed with the R package “survival” employing the Kaplan-Meier method.

A total of 5,946 participants who completed both ultrasound examination and H. pylori antibody testing were initially screened for this study [Figure 1]. Following the exclusion of 413 participants lacking follow-up data and 460 with missing baseline clinical characteristics, 5,073 individuals were included in the final analytical cohort. Of these, 2,630 (51.8%) were male, with a mean age of 43.3 ± 15.9 years. Hypertension and diabetes were present in 2,378 (46.9%) and 685 (13.5%) participants, respectively. A total of 2,394 (47.2%) participants were seropositive for H. pylori antibodies. Among the population, 1,610 cases met the MAFLD diagnostic criteria, with a prevalence of 31.7%.

Based on H. pylori antibody test results, the participants were categorized as seropositive or seronegative. Comparative analysis of baseline characteristics [Table 1] revealed that the seropositive group was significantly older (46.2 ± 16.0 years vs. 40.8 ± 15.4 years; P < 0.001) and had a greater percentage of male participants (55.3% vs. 48.8%; P < 0.001). Furthermore, the seropositive group exhibited a higher prevalence of diabetes, hypertension, and markers of metabolic dysregulation, including elevated BMI, waist circumference, FPG, HbA1c, and serum cholesterol and triglycerides. Indicators of liver injury (ALT and AST) were also elevated more frequently in the seropositive group. Notably, the prevalence of MAFLD was significantly higher among seropositive individuals than among their seronegative counterparts (36.6% vs. 27.4%).

As shown in Table 2, univariate logistic regression revealed that H. pylori seropositivity was a significant risk factor for MAFLD [unadjusted OR = 1.535, 95% confidence interval (CI): 1.363-1.729, P < 0.001]. This association persisted in multivariable analyses accounting for prespecified confounders. In Model 1 (adjusted for sex, age, race, diabetes, and hypertension), the OR was 1.422 (95%CI: 1.260-1.605, P < 0.001). After further adjustment for BMI, Platelets (PLT), CRP, FPG, serum cholesterol, ALT, and creatinine levels (Model 2), the OR remained significant at 1.282 (95%CI: 1.126-1.459, P < 0.001). These results confirm that H. pylori seropositivity is an independent predictor of MAFLD risk.

With a median follow-up of 28.8 years (interquartile range: 21.7-29.9), 1,891 deaths occurred within the cohort, resulting in an overall cumulative mortality rate of 37.3%. As shown in Table 1, individuals seropositive for H. pylori had significantly higher mortality rates than seronegative individuals (43.5% vs. 31.7%, P < 0.001). A significant difference in survival probability between the H. pylori-positive and negative groups was demonstrated by Kaplan-Meier analysis (log-rank P < 0.001; Figure 2A). This survival disadvantage associated with H. pylori seropositivity became apparent within the initial ten years of follow-up and continued to increase over time. By the 30-year mark, the absolute difference in survival was 11.8% (seropositive: 56.5% vs. seronegative: 68.3%). Consistently, the Kaplan-Meier curves showed significantly greater survival probabilities in H. pylori-negative individuals, a pattern evident in both the non-MAFLD [Figure 2B] and MAFLD [Figure 2C] subgroups (log-rank P < 0.001 for each).

To assess the relationship between H. pylori seropositivity and all-cause mortality within the entire cohort, Cox proportional hazards regression analysis was conducted. The findings are presented in Table 3 and Figure 3. The initial unadjusted analysis indicated a significant association between seropositivity for H. pylori antibodies and elevated all-cause mortality risk (HR = 1.525, 95%CI: 1.393-1.670). To further investigate this potential association and account for confounding factors, two multivariable models were constructed. Model 1 was adjusted for sex, age, race, diabetes, and hypertension. Model 2 included all variables in Model 1, with additional adjustments for BMI, PLT, CRP, FPG, serum cholesterol, ALT, and creatinine levels. The positive association between H. pylori seropositivity and all-cause mortality persisted in both adjusted models (Model 1: HR = 1.236, 95%CI: 1.123-1.360, P < 0.001; Model 2: HR = 1.253, 95%CI: 1.139-1.379, P < 0.001). This suggests that the observed relationship is independent of the evaluated confounders. Notably, this significant association persisted in the analyses stratified by MAFLD and non-MAFLD subgroups.

The results are presented for the total population, the non-MAFLD group, and the MAFLD group across three adjustment models: crude (unadjusted), Model 1 (adjusted for sex, age, race, diabetes, and hypertension), and Model 2 (additionally adjusted for BMI, PLT, CRP, FPG, serum cholesterol, ALT, and creatinine levels).

In summary, this population-based epidemiological follow-up study demonstrated a significant correlation between H. pylori seropositivity and an increased risk of MAFLD and all-cause mortality. These findings highlight the enduring public health impact of H. pylori as a chronic bacterial infection, indicating the need for further research to ascertain whether eradication of H. pylori could reduce the risk of MAFLD and mortality in the general population.